The late Mark Savage had written his representative in Congress about this matter back in '92. In the reply it is stated that the current processors (they've been upgraded since) could not handle all the radar data without bogging down. Since that time two major hardware upgrades have taken place, HOCSR & DSR (see glossary). Can the newer processors handle all this radar data yet?
(Original on file)
U.S Department
of Transportation
Federal Aviation
Administration
Office of the Administrator
800 Independence Ave., S.W.
Washington, D.C. 20591
FEB 13, 1992
The Honorable Chalmers P. Wylie
House of Representatives
Washington, DC 20515
Dear Mr. Wylie:
Thank you for your letter to former Deputy Secretary Elaine Chao on behalf of your constituent Mr. Mark A. Savage concerning radar data processing in the National Airspace System.
Extensive time has been devoted to the analysis of Mr. Savage's inquiries and the papers he mentioned that were prepared by Mr. Thomas G. Lusch and entitled "Real Targets-Unreal Displays: The Inadvertent Suppression of Critical Radar data." Our initial review was completed in 1989 when Mr. Lusch's longer paper on the same subject was forwarded to us from the Federal Aviation Administration's Great Lakes regional office. We have completed a detailed analysis of Mr. Lusch's recent submission. Additionally, we tasked Computer Sciences Corporation (CSC) to aid in the evaluation and analysis of the report. The CSC is a company which has been involved with operational software development for the en route system since 1976 and has developed a level of expertise unmatched in industry.
The results of the efforts of both offices lead us to the following conclusions regarding the concerns of Messrs. Savage and Lusch and the three recommendations presented by Mr. Lusch:
1. The recommendation to make the optional software modification (patch ZC150) mandatory for all operational sites.
We believe that each air route traffic control center (ARTCC) is in the best position to know what operational environmental configuration is most appropriate for its needs in this matter. This modification is available to all sites, and it is at the discretion of the facilities to utilize the modification. It is possible that some facilities (including the Cleveland ARTCC) may not have the conditions described as the cone-of-silence. Some facilities have determined that the disadvantages, such as excessive clutter on the controller display, outweigh the benefits of the patch. Consequently, we have recommended that this patch continue as a national local patch and remain optional for use by each ARTCC.
2. The recommendation that the scope of the modification (patch ZC150) be broadened to provide stratification and double preferred coverage for all radar sort boxes.
Implementation of this recommendation would require an excessive amount of the considerable computer resources available in the Host Computer System (HCS). The current compute element utilization (CPU) resources at an ARTCC at peak workload are approximately 35 percent. The CSC analysis of this approach determined that an additional 30 percent of CPU would be needed to process all radar data via this method. We do not know the exact additional CPU requirements for the integration of the upcoming Initial Sector Suite System (ISSS) but are unwilling to compromise the efficiency of the ISSS by bringing the existing computer load to 65 percent, even prior to ISSS implementation.
In addition, since the current processor is a uniprocessor, only one function is executed at any one instant. If the system is required to operate at peak capacity routinely, it is possible that the processing of data or instructions will be delayed to the degree that data will be lost or could affect the flight plan and target data bases, which could severely impact the reliability of the en route system.
Finally, there is some concern about the cost of redesigning and implementing such a change. It has been estimated that this developmental effort would take 30 person-years and about $4 million to accomplish. While these costs are significant, our major objection to this recommendation is the estimated computer resources that would be required to accomplish this change.
3. The recommendation to use all radar data and track all aircraft.
We agree with the benefit of this recommendation, and one of the goals for the Advanced Automation System project is to accomplish the same objective. However, it is impossible to accomplish this goal with the current hardware and software configuration.
In summary, Messrs. Savage and Lusch have described situation in the current en route system, including areas where radar coverage is not available or displayed. There are software methods to alleviate some of them. We continue to strive toward the ultimate goal of tracking all targets, utilizing all data acquisition information, and displaying them in an appropriate manner. An interim solution to the issues cited in their correspondence will be the ability to expand the number of radar sites interfaced with the HCS. We plan to incorporate this capability into the software system release scheduled for June 1992.
We thank you for your interest in this matter an hope this explanation will enable you to respond to Mr. Savage.
Sincerely,
Barry Lambert Harris
Acting Administrator
© 2000 by Lusch's Midair Collision Investigations. All rights reserved.